[all-commits] [llvm/llvm-project] c30ab6: [mlir] Transform scf.parallel to scf.for + async.e...

[Eugene Zhulenev via All-commits]

  Branch: refs/heads/master
  Home:   https://github.com/llvm/llvm-project
  Commit: c30ab6c2a307cfdce8323ed94c3d70eb2d26bc14
      https://github.com/llvm/llvm-project/commit/c30ab6c2a307cfdce8323ed94c3d70eb2d26bc14
  Author: Eugene Zhulenev <ezhulenev at google.com>
  Date:   2020-11-13 (Fri, 13 Nov 2020)

  Changed paths:
    M mlir/include/mlir/Dialect/Async/CMakeLists.txt
    M mlir/include/mlir/Dialect/Async/IR/Async.h
    M mlir/include/mlir/Dialect/Async/IR/AsyncBase.td
    M mlir/include/mlir/Dialect/Async/IR/AsyncOps.td
    A mlir/include/mlir/Dialect/Async/Passes.h
    A mlir/include/mlir/Dialect/Async/Passes.td
    M mlir/include/mlir/ExecutionEngine/AsyncRuntime.h
    M mlir/include/mlir/InitAllPasses.h
    A mlir/integration_test/Dialect/Async/CPU/lit.local.cfg
    A mlir/integration_test/Dialect/Async/CPU/test-async-parallel-for-1d.mlir
    A mlir/integration_test/Dialect/Async/CPU/test-async-parallel-for-2d.mlir
    M mlir/lib/Conversion/AsyncToLLVM/AsyncToLLVM.cpp
    M mlir/lib/Dialect/Async/CMakeLists.txt
    M mlir/lib/Dialect/Async/IR/Async.cpp
    A mlir/lib/Dialect/Async/Transforms/AsyncParallelFor.cpp
    A mlir/lib/Dialect/Async/Transforms/CMakeLists.txt
    A mlir/lib/Dialect/Async/Transforms/PassDetail.h
    M mlir/lib/ExecutionEngine/AsyncRuntime.cpp
    M mlir/test/Conversion/AsyncToLLVM/convert-to-llvm.mlir
    A mlir/test/Dialect/Async/async-parallel-for.mlir
    M mlir/test/Dialect/Async/ops.mlir
    A mlir/test/mlir-cpu-runner/async-group.mlir

  Log Message:
  -----------
  [mlir] Transform scf.parallel to scf.for + async.execute

Depends On D89958

1. Adds `async.group`/`async.awaitall` to group together multiple async tokens/values
2. Rewrite scf.parallel operation into multiple concurrent async.execute operations over non overlapping subranges of the original loop.

Example:

```
   scf.for (%i, %j) = (%lbi, %lbj) to (%ubi, %ubj) step (%si, %sj) {
     "do_some_compute"(%i, %j): () -> ()
   }
```

Converted to:

```
   %c0 = constant 0 : index
   %c1 = constant 1 : index

   // Compute blocks sizes for each induction variable.
   %num_blocks_i = ... : index
   %num_blocks_j = ... : index
   %block_size_i = ... : index
   %block_size_j = ... : index

   // Create an async group to track async execute ops.
   %group = async.create_group

   scf.for %bi = %c0 to %num_blocks_i step %c1 {
     %block_start_i = ... : index
     %block_end_i   = ... : index

     scf.for %bj = %c0 t0 %num_blocks_j step %c1 {
       %block_start_j = ... : index
       %block_end_j   = ... : index

       // Execute the body of original parallel operation for the current
       // block.
       %token = async.execute {
         scf.for %i = %block_start_i to %block_end_i step %si {
           scf.for %j = %block_start_j to %block_end_j step %sj {
             "do_some_compute"(%i, %j): () -> ()
           }
         }
       }

       // Add produced async token to the group.
       async.add_to_group %token, %group
     }
   }

   // Await completion of all async.execute operations.
   async.await_all %group
```
In this example outer loop launches inner block level loops as separate async
execute operations which will be executed concurrently.

At the end it waits for the completiom of all async execute operations.

Reviewed By: ftynse, mehdi_amini

Differential Revision: https://reviews.llvm.org/D89963